Whole Exome Sequencing of FLT3-ITD Sorafenib-Resistant Acute Myeloid Leukaemia

Acute myeloid leukaemia (AML) is a heterogeneous disease characterised by an abnormal increase in myeloblasts. Current treatment protocol employs the standard "7+3" chemotherapy regimen complemented with haematopoietic stem cell transplantation. However overall cure rate remains at 30-40%. Targeted therapy, especially the use of tyrosine kinase inhibitors, has been the main focus in the pursuit of an alternative solution. Internal tandem duplication at the juxatmembrane domain of fms-like tyrosine kinase 3 (FLT3-ITD) was found in around 30% of AML cases, conferring unfavourable treatment outcome and prognosis. Sorafenib, and other FLT3 inhibitors, successfully induced remission or partial remission in most of the relapsed/refractory FLT3-ITD AML patients (CR/CRi/nCR = 82.4%, n=34 in our sorafenib monotherapy cohort). Although consolidation with combination of sorafenib and chemotherapy or hypomethylating agents was reported to significantly extend the remission period, all patients relapsed with drug resistance. Mutations at the tyrosine kinase domain (TKD) were reported to alter the binding affinity of the drugs to the receptor and hence generate resistance. However, they were found only in around 25% of the patients while other non-mutational mechanisms of resistance have also been reported. Their relative importance in patients remains uncertain. We hypothesised that there is an emergence of drug-resistant clones acquiring new mutations or pathways to develop the resistance.

We have performed whole exome sequencing (HiSeq pair-end sequencing of 101 bp) for sorafenib-sensitive and sorafenib-resistant samples obtained from 8 FLT3-ITD relapsed/refractory AML patients before sorafenib treatment and at relapse with drug resistance. Genomic content of the bulk population was observed to change drastically among the 8 patients, with disappearance of single nucleotide polymorphisms (SNPs) and emergence of new SNPs, suggesting the emergence of drug-resistant clones. Emergence of TKD D835Y mutation was observed in 3 patients with variant allele frequencies (VAF) at 11%, 44% and 50% that were undetectable in the sorafenib-sensitive samples. Mutations at other reported sites of TKD were not found. Filtering of the generated list of single nucleotide changes and insertion/deletions (indels) to sort out emerging non-synonymous mutations with high VAF at resistant samples but undetectable in the sensitive samples was done. They include SLC15A1, CDC27, NADKD1, DGAT2, UBXN11, TFR2 and TTBK1, potentially implicated in sorafenib resistance. They are involved in different pathways, including membrane transporter, cell cycle regulation and energy metabolism. Further functional validation of their role in sorafenib resistance is required to understand the mechanisms involved.